International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 5

Number 1

March 2011

1 INSTRUCTIONS

A ship’s navigation instrument produces and pro-

vides a variety of information necessary for safe

navigation.

A number of researches’ efforts to provide more

advanced information for safe navigation.

Related studies include a study on collision

avoidance assistant system using fuzzy case-based

reasoning[1] and a study on a conceptual model for

collision avoidance for ontology-based fuzzy CBR

support system[2], and a study on a system that au-

tomatically sets up routes and supports the descrip-

tion of the set route for mates on small ships, who

are lacking of expertise[3]

In addition, there are a study that built an embed-

ded system that can control ships steering system of

ships with speech wheel order[4], speech-

recognition-based intelligent ships steering control

system suggested by Gyei-Kark Park and Ki-Yeol

Seo[5], and a study on ontology based fuzzy ships

steering control system[6].

However, studies so far have been carried on sys-

tems for navigators without expertise on small ships

without automatic control, and related systems or

studies which can provide comprehensive infor-

mation efficiently for experts are lacking.

This study tries to suggest data fusion models

necessary for the construction of the system to un-

derstand, unite, and provide the multimedia marine

information for mates. A ship’s navigation instru-

ment produces and provides a variety of information

necessary for safe navigation.

2 OVERVIEW OF A DATA FUSION MODEL

FOR NAVIGATION INFORMATION

Such a process as Figure 1 is necessary for establish-

ing a data fusion model to recognize a situation us-

ing the multimedia navigation safety information

provided by diverse navigation equipment, induce

the information needed for decision-making, and

provide it linguistically.

First of all, this paper composes a data field by

navigation equipment by analyzing the raw data

produced and provided by navigation equipment

such as GPS, ARPA, AIS, NAVTEX, VHF, etc. and

then establishes a knowledge representation model

to express the data in each data field and the rela-

tionship between objects as subjects between attrib-

utes linguistically using a semantic network.

Next, this paper establishes a data fusion model

using a knowledge representation model, and pro-

vides the information obtained newly by providing

the information of a data fusion model in a language

On a Data Fusion Model of the Navigation and

Communication Systems of a Ship

Gyei-Kark, Park & Young-Ki, Kim

Division of Maritime Transportation System. Mokpo National Maritime University,

Republic of Korea

ABSTRACT: Ship mates should be aware of images, numerical values, texts and audio-based information of

radar, AIS, NAVTEX, VHF, and etc. for safe navigation. However, it is very complicated and difficult for

them to acquire such information and use it as data for decision-making for safe navigation while keeping

watch for navigation. So, a system to understand, unite and provide multimedia marine information for mates

in voice is necessary. This study tries to suggest data fusion model of the navigation and communication sys-

tem.

or by fusing or inducing the information provided by

a data fusion model in a language.

Figure 1. Diagram of providing the inferred information

3 KNOWLEDGE REPRESENTATION AND

DATA FUSION MODEL

3.1 Data field of navigational equipments

To use the information provided by navigation

equipments, the information provided by the equip-

ment was analyzed and data field for the equipment

were prepared.

GPS, ARPA and AIS information should provide

which is specified in SOLAS resolutions.

NAVTEX provides information received in the

unit of a character of English alphabet in a text form,

natural language processing is completed so it was

assumed that an appropriate data field was obtained.

Most navigation warning, weather warning and

other urgent safety-related notices NAVTEX pro-

vides were discovery of a new navigation obstruc-

tion, change of aids to navigation, construction or

training section, or threatening weather occurrence,

etc., and such information consists of the name, the

term of validity and location or area.

VHF is communication equipment using frequen-

cy. Since it communicates using human language di-

rectly, information and types obtained using this

equipment are unlimited. However, the most im-

portant thing is that it has a merit that through which

the intention of the other ship can be known.

Using these, data fields were prepared.

To use information that can be provided by navi-

gational equipments, it was assumed that speech

recognition and natural language processing had

been completed.

3.2 Knowledge Representation Models

The Knowledge Representation Model by navigation

instruments was built by expressing the knowledge

relationship between the data related to the subject

and the subject using semantic network, and the in-

formation which each equipment provides was ex-

pressed in simple sentences, and the knowledge rep-

resentation by the subjects was constructed using the

knowledge representation by navigation instrument.

3.2.1 Knowledge Representation Model of GPS

All information provided by GPS is included.

Figure 2 is Knowledge Representation Model of

GPS and it expressed information which is provided

by this model in a simple sentence.

“Ownship’s Position’s GPS position Is latitude

34˚ 12.5˙N, longitude 126˚ 22.4˙E.”

“Ownship’s Speed Is 18.3 kts."

Figure 2. Knowledge Representation Model of GPS

3.2.2 Knowledge Representation Model of ARPA

All the information provided by ARPA is includ-

ed and D1: Bearing & D2: Range as location infor-

mation added vertices as position before connecting

with subject for amalgamation with information

which is provided by other navigation instrument.

Figure 3 is Knowledge Representation Model of

ARPA, and it expressed information which is pro-

vided by this model in a simple sentence.

“Ship 1’s Position is Bearing 312 degree.”

“Object 1’s CPA Is 0.0 miles.”

Figure 3. Knowledge Representation Model of ARPA

3.2.3 Knowledge Representation Model of AIS

All information which is provided by AIS can be

expressed but only D14: GPS position as location in-

formation was connected to ship as subject through

the vertices as position.

Figure 4 is Knowledge Representation Model of

AIS, and it expressed information which is provided

by this model in a simple sentence.

“Ship 1’s Heading Is 000 degree.”

“Ship 1’s Name is SAEYUDAL.”

Figure 1. Knowledge Representation Model of AIS

3.2.4 Knowledge Representation Model of NAVTEX

It was designed using data field of NAVTEX.

Figure 5 is Knowledge Representation Model of

NAVTEX, and it expressed information which is

provided by this model in a simple sentence.

“Object 1’s Name is Dangerous wreck.”

“Object 1’s Position is latitude 34˚ 12.5˙ N and

longitude 126˚ 22.5˙ E.”

Figure 5. Knowledge Representation Model of NAVTEX

3.2.5 Knowledge Representation Model of VHF

It was designed using data field of VHF.

Figure 6 is Knowledge Representation Model of

VHF, and it expressed information which is provid-

ed by this model in a simple sentence.

“Ship 1’s Intention Is overtake.”

Figure 6. Knowledge Representation Model of VHF

3.3 Data fusion algorithm

A data fusion process is necessary for fusing a

knowledge representation model into a data fusion

model.

In order to judge that the data provided by two

knowledge representation models are the infor-

mation of the same objects, the data with the same

meaning among the data provided by two knowledge

representation models should be comparable, and it

can be judged that the data provided by two

knowledge representation models are the infor-

mation with the same objects when their similarity is

within a certain range after comparison.

Figure 7 shows a proposed data fusion algorism.

Figure 7. Data fusion algorithm

3.4 Data fusion model

3.4.1 Data fusion model in the case of a ship as

target

In the case of ship as subject, information can be

obtained using ARPA, AIS and VHF in general.

Therefore, it was represented by combining

knowledge representation models of ARPA, AIS and

VHF.

Figure 8 is data fusion model in the case of a ship

as target.

Figure 8. Data fusion model in the case of a ship as target

3.4.2 Data fusion model in the case of a object as

target

In the case of subject as except ship, the infor-

mation can be obtained by ARPA or NAVTEX etc,

so it was represented by combining knowledge rep-

resentation models of ARPA and NAVTEX.

Figure 9 is Data fusion model in the case of a ob-

ject as target.

Figure 9. Data fusion model in the case of the subject as except

ship

4 APPLICATION OF DATA FUSION MODEL

4.1 Description of navigation situation using data

fusion model by subject

It explained the navigation situation and the infor-

mation of other ships and marine obstacle which ma-

te can obtain in the given navigation situation was

expressed using Data Fusion Model by subject in the

sentence.

The navigation situation of Figure 10 is a danger-

ous one in which two ships are encountering while

they pass through a narrow channel.

Figure 10. Navigational Situation “Dangerous Stage”

4.2 Navigation situation expression of information

that can be obtained from Data Fusion Model

Figure 11 shows a data fusion model for the infor-

mation acquirable in a dangerous situation of scenar-

io.

The data fusion model in Figure 11 can provide

all the information provided by navigation equip-

ment in a navigation of scenario in a simple sen-

tence.

This expresses the meaning of information on

Ship1, Ship2 and Object1 provided by a given data

fusion model in a simple sentence.

“Ship 1’s Position Is D1:Bearing Is 304˚.”

“Ship 1’s Position Is D2:Range Is 2.12 miles.”

“Ship 1’s Position Is D14:GPS position Is Lati-

tude 34˚ 19.4˙ N and longitude 126˚ 05.85˙ E.

“Ship 1’s D3:CPA Is 0.15 miles.”

“Ship 1’s D4:TCPA Is 5.2 minute.”

“Ship 2’s D5:True course Is 126˚.”

“Ship 2’s D6:Ture speed Is 10.0 kts.”

“Ship 2’s D7:IMO No. Is 440100001.”

“Ship 2’s D18:Rate of turn Is 0.0˚/min.”

“Object 1’s Position Is D1:Bearing Is 332˚.”

“Object 1’s Position Is D2:Range Is 1.7 miles.”

“Object 1’s Position Is D14:GPS position Is Lati-

tude 34˚ 19.75˙ N and longitude 126˚ 07.0˙ E.

“Object 1’s D9:Name Is Dangerous wreck.”

“Object 1’s D23:Period Is Unlimited.”

Figure 11. Navigation situation expression using the Data Fu-

sion Model for each object

4.3 Navigation situation expression using the Data

fusion model

Figures 12, 13 and 14 express the information ob-

tainable through induction by a mate in a navigation

situation of scenario using a data fusion model in

Figure 11.

Figure 12 expresses the information of a new

meaning inducing sentence using the meaning of

multiple objects “the CPA of Ship1, Ship2 and Ob-

ject1 is all within 1mile and less than 7 minutes, so it

is very dangerous” using a data fusion model.

The meaning except for the meaning created new-

ly after induction can be expressed linguistically.

- Ship1`s D3, D4, Ship2`s D3, D4, Object1`s D3,

D4.

“Ship1's CPA Is 0.15mile, TCPA Is 5:2, Ship2`s

CPA Is 0.0mile, TCPA Is 4:17, Object1`s CPA Is

0.6mile, TCPA Is 6:24.”

Figure 12. Navigation situation expression – 1

Figure 13 expresses the information of a sentence

induced by combining the meaning of an object ac-

quired from multiple navigation equipment “Object1

has Bearing 332°, Range 1.7mile, CPA 0.6mile, and

TCPA 6:24, and its name is Dangerous wreck”using

a data fusion model.

All the meanings of a sentence can be expressed.

-Object1`s D1, D2, D3, D4, D9.

“Object1 `s Bearing Is 332°, Range Is 1.7mile,

CPA Is 0.6mile, TCPAIs 6:24, Name Is Dangerous

wreck.”

Figure 13. Navigation situation expression - 2

Figure 14 expresses the information of a sentence

inducing a situation and presenting a solution by

combining the meanings of an object acquired from

multiple navigation equipment “Ship 1 has CPA

0.15mile and TCPA 5min 2sec, but it doesn’t veer

and respond to communication, so do DSC(Digital

Selective Calling) for IMO No. 440100002”using a

data fusion model.

The induced meaning cannot be expressed, but

the information can be expressed.

- Ship1`s D3, D4, D18, D25, D7.

“Ship1`s CPA Is 0.15mile, TCPA Is 5min 2sec,

Rate of turn Is 0.0°/min, Intention is nothing, DSC Is

440100002.”

Ship1`s CPA Is 0.15mile, TCPA Is 5min 2sec,

But, Rate of turn Is 0.0°/min, Intention is nothing ∴

DSC 440100002.

Ship1`s D3, D4 But D18, D25, DSC D7.

Figure 14. Navigation situation expression - 3

5 CONCLUSIONS

This study proposed a data fusion model using se-

mantic network to analyze the information provided

by each navigation equipment and express the mean-

ing of information provided by navigation equip-

ment by selecting navigation equipment such as

GPS, ARPA, AIS, NAVTEX, VHF receiver, etc.

providing essential information for grasping a navi-

gation situation, and explained the information ac-

quirable and the inducible by a mate in a navigation

situation of scenario using a data fusion model.

The test of the proposed model in some real navi-

gation situations will be done to verify its validity in

future.

REFERENCES

Gyei-Kark Park, John Leslie Benedictos, “Ship Collision

Avoidance Support System Using Fuzzy-CBR”, Korean In-

stitute of Intelligent System, Journal of Korean Institute of

Intelligent System, Vol. 16, No. 5, pp. 635-641, 2006.

Gyei-Kark Park, Woong-Gyu Kim, John Leslie RM Benedic-

tos, “Conceptual Model for Fuzzy-CBR Support System for

Collision Avoidance at Sea Using Ontology”, Korean Insti-

tute of Intelligent System, Journal of Korean Institute of In-

telligent System, Vol. 17, No.3, pp. 390-396, 2007.

Tae-ho Hong, Ki-Yeol Seo, Gyei-Kark Park, “Building of an

Integrated Navigation Guiding System Based on ENC”,

Korean Institute of Intelligent System, Proceedings of KIIS

Spring Conference, Vol. 15, No.1, pp. 394-399, 2005.

Gyei-Kark Park, Ki-Yeol Seo, Tae-ho Hong, “Development of

an Embedded System for Ship’s Steering Gear using Voice

Recognition Module”, Korean Institute of Intelligent Sys-

tem, Journal of Korean Institute of Intelligent System,

Vol.14, No.5, pp.604-609, 2004.

Ki-Yeol Seo, Se-Woong Oh, Sang-Hyun Suh, Gyei-Kark Park,

“Intelligent Steering Control System Based On Voice In-

structions”, International Journal of Control Automation

and System, Vol. 5, No.5, 10, pp. 539-546, 2007.

Ki-Yeol Seo, Gyei-Kark Park, Chang-Shing Lee, Mei-Hui

Wang, “Ontology-based Fuzzy support agent for ship steer-

ing control”, Expert systems with applications, Vol 36, Is-

sue 1, pp. 755-765, 2009.